FRACTIONATION

Field of the Invention

[0001] The present invention relates in general to centrifugal microfluidic fractionation with off-chip pressure control, and in particular to a method, chip, system, and kit for density-based extraction of a component of a multi-component sample fluid.

Background of the Invention

[0002] Fluid sample analysis is a commercially important activity. Natural samples, in particular, are frequently highly complex and stable mixtures. Many biofluids (e.g. blood, milk, urine, saliva, tears, sweat, egg, sperm), and naturally found samples such as hydrocarbons, wastewater, effluent, runoff, and food and chemical products and biproducts are frequently aqueous liquids carrying a very wide range of dissolved, suspended or included particles, macromolecules, bodies, cells, organoids, and like masses. Furthermore, a very wide variety of oil- or water-based samples are produced to study solid, liquid or gaseous material, the samples being produced to have fluid properties that facilitate transport and control fortesting.

[0003] Microfluidics is an emergent field offering great advantages in multiplex analysis of small liquid volumes. By moving tiny volumes of fluids around on low cost plastic chips, a variety of testing and assaying procedures can be produced. By centrifuging, while providing a pressure controlled supply at respective ports of the chip under centrifuge, a large variety of automated microfluidic procedures can be provisioned. For example, according to Applicant’s WO 2015/132743, the following are efficient ways of controlling pressure at a port of a centrifugal microfluidic chip (i.e. off-chip) during centrifugation: pneumatic slip rings (rotary couplings); centrifuge-mounted pressurized fluid supplies; centrifuge-mounted pumps; and a centrifuge-mounted electronic valve for opening a closed microfluidic path to ambience. Herein a “port” is understood to be a fluidic interface of a microfluidic chip, which can be used for various purposes, including as a vent, as a loading port, and as a controlled pressure supplied port. A port only used as a vent can be smaller than others. A vent used for manual loading may be larger than others, and a loading port or controlled pressure supplied port may have features to assist in establishing a sealed connection, such as a luer lock or other standard coupler.

[0004] Refinement of fluid samples may be performed with various reaction-based, membrane- and filter-based, and separation techniques, and a large body of the study is devoted to this, but fractionation (herein referring only to mass-density fractionation, as opposed to crystallization, distillation, solubilization, or other fractionation techniques) has many advantages over other candidates, in that: it has low energy demands, especially in comparison with techniques calling for phase changes; the sample is not denatured or altered by any extreme temperature change; it typically has no effect on chemical or biological constitution of the sample; it does not rely on supply, chemical purity, or state of, reaction compounds; it has little sensitivity to temperature, pressure or environmental conditions; it exposes the sample to very low risk of cross-contamination; and it offers a very simple, physically reliable, differentiating mechanism. The only limits to practical application are that a desired fraction of the sample have a (mass) density that is narrowly enough constrained, and differentiated from other components of the sample.

[0005] Fractionation is of greater value for purification, separation, or quantification, the more components the sample contains, or may contain, assuming a reasonable variability of mass densities of the desired components. One step isolation of a component from a complex mixture is possible to the extent that the component has a unique range of densities, and nothing else in the mixture has a density within this range, though this is not always feasible with most natural or complex samples. The narrower the range, the more precise the extraction point must be defined. While tuned density media can be added to greatly facilitate the extraction point, this is only possible when the desired component is characterized. Furthermore, the use of density media of a particular colour, clarity or other discernible feature, demarcates the sample in centrifuge vials, but does not appreciably assist in automation, and has limited demonstrated use in fractionation in centrifugal microfluidics.

[0006] To keep the costs of microfluidic chips low enough for single use applications, and thereby avoid issues with cleaning and contamination, it is advantageous to avoid multi-material inserts and electronics (especially those that are exposed to sample fluids), and this generally leaves one with fixed microfluidic channels and chambers. The control of pressure supplies at ports of the chip enables a variety of processes without increasing chip costs or complexity but system costs increase with the number of independently (pressure) controlled ports.

[0007] Fixed volumes in reservoirs and fixed extraction points on fractionation columns are particularly problematic in view of variabilities of component volumes within many, especially natural, liquid samples. Intrinsic variability from one sample to another poses a challenge for automated fractionation in centrifugal microfluidics. While precise metering (for example using Applicant’s issued patent family: WO2013003935) is known in the art, and this can ensure an initial volume within the fractionation chamber within narrow limits, but it cannot account for intrinsic variability of content fractions, such as a volume of plasma in whole blood. This leads to less selective extraction of a desired component (with more undesired components) or less than complete extraction of the desired component.

[0008] US 6,719,682 to Kellogg et al. teaches, with regards to FIGs. 9A-H, a centrifugal chip for fractionation. It specifically uses a U shaped channel with an overflow chamber 404 that allows for imprecise loading of the sample in chamber 401 . As long as the sample introduced exceeds a volume (and not by more than the overflow volume), this chip layout ensures that a fixed volume will be in the separation column 403. However, as mentioned above, this system does not account for sample composition variability. While the extraction capillary 406 may be well aligned for one sample of blood, another sample will have more red blood cells, and less plasma, resulting in a misalignment of the extraction capillary 406. For many fluid samples there is generally no way of knowing the volume of each component until after centrifugation. Thus without testing, one can only choose a location for capillary 406 based on statistics, and accept error in terms of impurities or incomplete extraction.

[0009] While in theory, one could introduce a metered volume of density media into the column 403 with the blood sample, and this will serve to increase a spatial separation between the distinct phases of the blood sample, it still cannot so much address the issue, as ensure that it if the desired component is not collected, it will be the density media that will be included in the extracted liquid, as opposed to other components of the blood. While this could look like an acceptable compromise, a sample with a different volumetric composition than prescribed for a chip, may end up only extracting the media, which would generally be useless. The “buffy coat”, for example, is typically a very small volume fraction of the blood, and is easy to miss.

[0010] For example, a paper to Scott T. Moen et al. entitled A Centrifugal Microfluidic Platform that Separates Whole Blood Samples into Multiple Removable Fractions Due to Several Discrete but Continuous Density Gradient Sections (PLoS ONE 11 (4): e0153137. https://doi.org/10.1371/journal.pone.0153137), teaches the use of different density media in respective “lanes” of their separating column. Applicant notes the introduction, which states: “density gradient centrifugation process requires trained personnel and a fair amount of dexterity to load density layers and accurately extract the desired blood fractions. In addition, a relatively large amount of blood is needed (typically 1-10 ml_ [8][9]) to observe the discrete band of leukocytes for extraction. However, in most clinical applications, it would be advantageous to have the option of using smaller amounts of blood to perform the analysis.” [0011] Another technique that does not address, but can improve accuracy of extraction, is the division of the column into axially proximal and distal chambers separated by a channel, or otherwise pinching a waist of the separation column adjacent the extraction outlet. This allows for greater precision of extraction, when operated by a human, somewhat like the use of a density medium, but still does not ensure that the extraction outlet is well aligned with a desired component.

[0012] WO 2014/111721 to Banks shows separation chambers 230,235 interconnected by separation channel 231 , which is said to introduce a “pinch point” between the first and second separation chambers. Banks notes that a single chamber with a narrow portion would work as well. Banks states that the reason for the pinch-point is to reduce “remixing of the fluid within the separation chamber(s) after separation”. It appears to Applicant that by narrowing the separation column, a location of the extraction channel can be more precisely aligned with a fractionated sample volume, as an area of an interface between separated phases is reduced. With correct alignment to extract all of, or exclusively, the desired component, substantially less of the other components/density medium, or substantially more of the desired component is extracted.

[0013] Applicant notes that fractionation of whole blood is an important process in many clinical applications such as cancer diagnosis, [1] autoimmune disease[2] and biomedical research. [3] In general, whole blood fractionation is conducted in a laboratory setting by carefully layering the blood sample on a density gradient medium followed by centrifugation. [4] This method is the most frequently used approach in the lab for blood handling, along with other sample preparation techniques, such as erythrocyte lysis[5] and magnetic separation. [6] However, the sample handling process involves complex procedures and is often performed manually with multiple pipetting steps; it is a time consuming and labor intensive work that requires experienced and trained operators for reliable and reproducible results. An automated fractionation device and procedure is therefore highly desirable.

[0014] Accordingly there is a need for a microfluidic chip, a microfluidic system, and method for improved centrifugal microfluidic fractionation. The improvement may involve improved automation, time to fraction extraction, or accuracy (reduced undesired component inclusion in extraction, and/or increased desired component extraction).

Summary of the Invention

[0015] Applicant realized a low-cost, highly versatile, and robust microfluidic technique for fractionation, including novel microfluidic chips, kits, a system, and methods. The technique involves or allows controlling pressure supplied at ports of the chip to vary a level (radius from axis of centrifuge) of an interface between phases (isopycnic surface, which appears as a line in 2D renderings) relative to the chip. Specifically, this is accomplished by delivery of a medium into the fractionation column at a controlled rate and volume to displace the isopycnic surface. The isopycnic surface can be slowly displaced without stopping centrifugation, and without remixing or disrupting the fractionation. The use of one or more gradient medium prior to fractionation may assist in demarcating isopycnic surfaces of interest, and the same medium can be used for displacing the fractionated sample.

[0016] Accordingly a method is provided for centrifugal microfluidic fractionation, the method comprising: providing a fluid sample in a fractionating column of a centrifugal microfluidic chip; centrifuging to fractionate the sample; and, while centrifuging: operating a first off-chip pressure supply line coupled to a first port of the chip to dispense a volume of a first medium of characterized density into the column, until a desired fraction of the fractionated sample is aligned with an extraction channel that meets the column; and operating a second off-chip pressure supply line coupled to a second port of the chip to draw some of the fractionated sample located axis proximal of the extraction channel, into the extraction channel.

[0017] The first off-chip pressure supply line preferably pressurizes a medium chamber to a vented transit chamber located axis proximally of the extraction channel. A fluid dynamic resistance between the transit and medium chambers may be provided to ensure that the first medium is delivered to the transit chamber as a discretized droplet stream; or a fluid dynamic resistance between the transit chamber and column limit delivery of a metered volume only when the hydrodynamic resistance is overcome by the pressure in from the second pressurized fluid supply.

[0018] Also a method for centrifugal microfluidic fractionation is provided, the method comprising: mounting a centrifugal microfluidic chip controller with controllers for first and second pressurized fluid supplies, and a centrifugal microfluidic chip, on a centrifuge, for spinning the chip and at least part of the controller on an axis of the centrifuge at a rate of at least 5 Hz, the chip having a microfluidic network including: a vented fractionation column enclosing a volume the column extending between a proximal and a distal point relative to the centrifuge axis; an extraction channel meeting the column between the proximal and distal points, through which a component of a fractionated fluid can be removed from the column by operation of the first pressurized fluid supply; a vented transit chamber having an ingress and an egress; a medium chamber operably coupled to the second pressurized fluid supply containing a first medium of characterized density; a first and second conduit respectively coupling the medium chamber with the ingress, and the egress with the column; and a hydrodynamic resistive element between the column and medium chamber limit flow between these two chambers without applying a pressure at the second pressurized fluid supply; providing a fluid sample in the column; centrifuging a fluid sample in the column to fractionate the sample into a plurality of components having different mass densities; without stopping the centrifuge, dispensing a controlled volume of the medium into the column via the transit chamber, to displace the fractionated sample in the column until a desired fraction of the fractionated sample is aligned with the extraction channel; and operating the first pressurized fluid supply to selectively extract some of the fractionated sample via the extraction channel.

[0019] In either method, the medium may have a higher mass density than the sample, and if so is preferably injected into the column axis distally of the extraction channel; or the medium may have a lower mass density than the isopycnic surface, and is preferably injected into the column axis proximally of the extraction channel. The column may comprise a U shaped chamber where the axis distal point is a U bottom, whereby added medium on one branch of the U shaped chamber moves the fractionated sample into the other branch.

[0020] The column, whether U shaped or not, may be waisted in that the column is narrower at the extraction channel than away from the extraction channel. The chip may further comprise at least two medium chambers, or a mixing chamber for producing the first medium as well as one or more second density media having a different density than the first medium, and the method may further comprise supplying a volume of the second medium into the column prior to centrifugation.

[0021] Either method may further comprise a camera and illumination equipment for imaging the chip during centrifugation and software for analyzing the image data to determine the isopycnic surface relative to the extraction channel in the column, and controlling the pressures at the first and second pressurized fluid supplies in response to the analysis of the image data.

[0022] Providing the fluid sample may comprises loading the sample from a vial that is mounted to the chip controller; or extracting the part of the fractionated sample comprises ejecting withdrawn fluid to a vial that mounted to the chip controller.

[0023] Also accordingly, a centrifugal microfluidic chip for mounting to a centrifuge for rotation about an axis of the centrifuge is provided. The chip has a microfluidic network comprising: a fractionation column extending between a proximal point and a distal point relative to the axis, the proximal point being at, or coupled to, a first port; an extraction channel meeting the column between the proximal and distal points, through which fluid is removed from the column or blocked during centrifugation, depending on a pressure at a second port of the chip relative to the first port; a transit chamber having an ingress, and an egress coupled to the column, and a third port coupled thereto; a first medium chamber operably coupled to a fourth port of the chip and to the ingress, such that a change in pressure at the fourth port relative to a pressure in the transit chamber, can be used to draw fluid from the medium chamber into the transit chamber during centrifugation; and a hydrodynamic resistive element between the column and medium chamber that limits flow therebetween during centrifugation.

[0024] The column may: be waisted, in that the column is narrower at the extraction channel than at axis proximal and axis distal points on the column; or have a U shape with two branches extending axis proximally of a U bottom.

[0025] The chip may further comprise at least two medium chambers, or a mixing chamber coupled to two or more other medium chambers for producing the first medium as well as one or more second density media having a different density than the first medium; or the extraction channel forks, leading to two or more respective chambers, for respective fractionated components.

[0026] A volumetric capacity of the medium chamber may be at least half a volume of the column axis proximal of the extraction channel.

[0027] Furthermore, a centrifugal microfluidic system is provided comprising the chip loaded with fluids, for example with the medium chamber containing a medium having a density higher than the isopycnic surface of the intended fractionated sample, and the transit chamber is coupled to the column axis distally of the extraction channel; or the medium chamber comprises a medium having a density lower than the isopycnic surface of the intended fractionated sample, the transit chamber is coupled to the column axis proximally of the extraction channel, and the column comprises a U shaped chamber where the axis distal point is a U bottom, whereby added medium on one branch of the U shaped chamber moves the fractionated sample into the other branch.

[0028] The system mounted to a microfluidic chip controller is also provided, having a respective pressurized fluid supply lines coupled to the second and third ports, and respective flow controllers for controlling pressures thereat, the microfluidic chip control and chip being mountable to a centrifuge for rotation about an axis thereof.

[0029] The system may further comprise: a camera and illumination equipment for imaging the chip during centrifugation; a processor adapted to analyze the chip imaging to determine isopycnic surfaces relative to the extraction channel in the column; and controlling pressures at the first and second ports, in response to the analysis of the image data.

[0030] Finally, a kit is provided, the kit comprising the chip defined above, and a supply of at least one medium of a characterized density of a desired isopycnic surface of a desired component to be extracted from a sample. The kit may further comprise a centrifugal microfluidic chip controller with controllers for first and second pressurized fluid supplies, adapted for mounting the chip to a centrifuge, for spinning the chip and at least part of the controller on an axis of the centrifuge at a rate of at least 5 Hz.

[0031] A copy of the claims are incorporated herein by reference. Further features of the invention will be described or will become apparent in the course of the following detailed description.

Brief Description of the Drawings

[0032] In order that the invention may be more clearly understood, embodiments thereof will now be described in detail by way of example, with reference to the accompanying drawings, in which:

[0033] FIG. 1 is a top plan view of a centrifugal microfluidic fractionating chip, in accordance with a first embodiment of the present invention;

[0034] FIGs. 1A-E illustrate the chip of FIG. 1 in a sequence of states to perform a first centrifugal microfluidic fraction extraction process;

[0035] FIGs. 1C’,D’,E’ illustrate a first variant of the chip of FIG. 1 in a sequence of states performing essentially the same fraction extraction process, but with a displacing fluid having a density less than that of one component of the sample fluid, but greater than a density of an extracted fraction;

[0036] FIG. 2 is a top plan view of a centrifugal microfluidic fractionating chip, in accordance with a second variant in which the fractionation column is U shaped, and waisted about its exit, and the displacing fluid is fed axis-proximally, the view further comprising a set of levels within the column at respective stages of fraction-extraction; [0037] FIG. 2’ illustrates the second variant chip showing a set of levels within the column at respective stages of an alternative fraction-extraction process with an initial volume of a dense medium;

[0038] FIG. 3 is a top plan view of a single plane of a centrifugal microfluidic fractionating chip in accordance with a third variant, in which the fractionation column has an overflow chamber, the displacing fluid is fed axis-proximally, and the outlet consists of four axis-concentric through bores; the illustration further comprising a set of levels at respective stages of fraction-extraction;

[0039] FIGs. 4A-C illustrate a top plan view of a fourth variant of a centrifugal microfluidic fractionating chip, in respective, sequential, states of fraction-extraction, the fourth variant having a hydrodynamic resistance between the column and a metering chamber with an overflow;

[0040] FIG. 5 illustrates a top plan view of a fifth variant of a centrifugal microfluidic fractionating chip, the fifth variant has a column waisted about its exit and fed the displacing fluid axis-distally, the view further comprising a set of levels within the column at respective stages of the first fractionation extraction process implicating two gradient media resolving a selected range of mass densities;

[0041] FIG. 6 is a schematic illustration of a sixth variant of a centrifugal microfluidic fractionating chip further comprising four chambers for compounding a tailored density medium;

[0042] FIG. 7 is a schematic illustration of a seventh variant of a centrifugal microfluidic fractionating chip different from that of FIG. 1 in that a chamber is provided for retaining an extracted fraction; the chip is mounted to a centrifugal microfluidic controller which provides an on-board pressure supply integrated with a blade of a centrifuge;

[0043] FIG. 7 A is a schematic illustration of the seventh variant chip mounted via the chip controller to a centrifuge equipped for image-based controlled fraction-extraction; [0044] FIG. 8 is a schematic of an eighth variant centrifugal microfluidic fractionating chip built and used to demonstrate the present invention, and FIG. 8A is a photograph of the chip used to extract a buffy coat and plasma from blood;

[0045] FIGs. 9A-I are photographs showing a same column in different states of fractionation; and

[0046] FIGs. 10A,B are photographs of a same fractionation chamber in a same state, with and without colour filtering, showing how easily an isopycnic surface can be resolved for blood imaging. Description of Preferred Embodiments

[0047] Herein a technique for centrifugal microfluidic fractionation and extraction is taught. The technique involves providing centrifugal microfluidic chips with certain features, and providing pressure control at respective ports of the chips, to fractionate a liquid, and move a content of the column to align an isopycnic surface with an extraction channel, and extracting the component from the mixture. The chief advantages are the lower cost equipment, and ease of control, possibility of automation, small sample volume, and the higher selectivity and/or higher yield, of the extraction, especially if inherent compositional variability of the sample is modest or high, and the isopycnic surface can be identified by visible, optical, or index of refraction measurement.

[0048] FIG. 1 is a schematic top plan view of a centrifugal microfluidic chip in accordance with a first embodiment of the present invention. The chip is adapted to be mounted to a centrifuge and coupled with pressure controlled supplies of a centrifuge mounted pneumatic chip controller at respective ports. Operation of the centrifuge and the pressure controlled supplies permits fractionation of a sample and more selective, or higher yield, extraction of a desired part of the sample, with less equipment and workload than possible using prior art low volume microfluidic fractionation techniques.

[0049] As is common to all embodiments of the present invention, the chip includes a re lief- patterned surface, preferably of a low-cost single use, or possibly an easily and reliably cleaned multi-use substrate 10. The substrate 10 is covered by a covering layer (not in view) that may be an integral part of a cartridge, or may improve a stiffness of the chip making it more easily handled and manipulated. The substrate 10 and covering layer are preferably adherent to provide a bond with sealing properties, for example with one of the two composed of a thermoplastic elastomer as claimed in Applicant’s US 10,369,566. In particular, a TPE such as oil-free MedipreneTM and a rigid thermoplastic like ZeonorTM have been found to be very compatible materials and have a good sealforming bond. As such, the substrate 10, when sealed against the covering layer, encloses chambers within the chip, exposing the network only at ports 12. While the ports 12 may alternatively be provided at edges of the chip (see 39 of FIG. 3) as this would simplify chip construction, and avoid a need for through bores in either the substrate 10 or cover, they are more easily viewed and more often supplied with through holes in the covering layer or substrate as shown.

[0050] The through holes maybe provided in the substrate 10 as shown, or through the covering layer. There are advantages to providing through holes in an elastomeric material, such as TPE, whereby slight compression of the elastomeric material by a tube or other channel supporting body formed of a stiffer material can facilitate sealing.

[0051] The chip may be a single-substrate device, in that only the substrate bears any relief pattern, as shown. Alternatively, the chip may include a relief patterned substrate and a covering layer that has only through bore holes for ports, or the chip may comprise a single substrate relief patterned on both sides with one or more connecting through bores, with two covering layers on either side (ports either being provided on edge, or through one or more of the covering layers). Advantageously, multi-layered chips can be provided with alternating layers of TPE and TP for solvent-free bonding, with suitable alignment between the layers.

[0052] While FIG. 1 shows all of the chip features on a single side of a single substrate 10, it will be evident to one skilled in the art that it is trivial to place some of the numbered elements on different layers of a multi-layer chip to achieve the same effect. The only particular limitations on useful multi-layered chips for present purposes is that a monitored region 14 remain scrutable, preferably by illumination and most preferably in the visible spectrum. As such, 1- all layers from an inspection side of the chip to the region 14 are preferably highly transparent to inspection wavelengths, and free of any microfluidic channels that would impair inspection of the region 14 (in particular, the monitored region is preferably on a top or bottom substrate of any such stacked chip, as this will avoid multiple reflections of inspection wavelengths); and 2- the substrate or it’s adjacent layer (depending on orientation of the inspection wavelengths) is preferably opaque, and provides good contrast for imaging intended levels with respect to the sample, particularly the fractionated sample, and any coloured density medium.

[0053] Bonding the covering layer over the relief patterning encloses a network of microfluidic channels and chambers which define: a fractionation column 15, having an outlet 16 for withdrawing an extracted fraction from the column 15 when operated by a flow control mechanism; a medium chamber 17 for retaining a medium of a desired mass density; and a controlled delivery channel 18 between the medium chamber 17 and column 15. The controlled delivery channel 18, as illustrated, has a J shaped low flow impedance channel 19, a high flow impedance serpentine channel 21 , and a vented opening 20 connecting the two channels. The J channel 19 meets the column 15 axis- distally of the region 14 nearer the axis-distal end of the column 15.

[0054] The serpentine channel 21 extends from an axis-distal end of medium chamber 17 to the opening 20, which is axis-proximal of (any fill level of) both the medium chamber 17 and the column 15, as shown. This is beneficial for making it very difficult for any fluid, under any centrifugal protocol, to ever back flow from the J channel 19 into the serpentine channel 21 , or for an air plug located in the opening 20 to be displaced. As such, the opening 20 is never blocked. The hydrodynamic resistance of the serpentine channel 21 resists flow until and unless a threshold pressure difference between medium chamber 17 and opening 20 is sustained for a priming period, and cooperates with a nozzle 22 defined where it meets the opening 20 to ensure a droplet-based discretization of the medium, in use. The serpentine channel 21 meets the vented opening 20 at a nozzle 22. Nozzle 22 may preferably be designed for a variety of medium liquids to emit a droplet stream while under centrifugation at a given rate and port 12c (see FIG. 1A) is supplied with a positive pressure relative to port 12b. Droplet size and rate are typically controllable by varying the pressure and centrifugation rate.

[0055] Applicant has found that excellent metering can be provided with this structure, in that the droplets of small, regular, volume are dispensed at a very regular rate with a substantially uniform pressure applied at the port, and a substantially constant centrifugal force (within limits provided by standard equipment). As dispensing is quantized, in that droplets only fall once they’re of a certain volume, the dispensed volume can be quite well controlled with time. With readily available optical imaging, and machine vision based feedback, accurate control over displacement of a visible isopycnic surface within a fractionated sample can be automated. Droplet size and rate can also be varied by changing centrifugation rate, and/or pressure.

[0056] While axis-relative position of the opening 20 relative to medium chamber 17 can be made quite irrelevant (except to the extent that it influences the threshold pressure difference and period) with additional controls over ambient pressure in the opening 20, it is nonetheless effective for control to place the opening 20 axis proximal (any fill level within) the medium chamber 17. The axis-relative position of the opening 20 relative to the column 15 is another matter, and it has to be above the fill level of the column 15 to avoid bubble mixing in the J-shaped channel 19, and any possibility of sample fractions entering the serpentine channel 21 , in the illustrated embodiment.

[0057] The column’s outlet 16 communicates with a hole 12', which may be used as a port, or as a via to another layer of the chip. If a via, the other layer does not cover the ports 12 on the axis-proximal edge of substrate 10, or has through holes aligned therewith. If a via, a closed channel coupling to hole 12' either passes axis proximal of the fill level of column 15 (preferably any fill level within the column), or the channel opens to a vented chamber with some means of pressure control at the corresponding port: either the port is coupled to a pressurized fluid supply operable to control extraction through outlet 16, or the port is valved for selective opening to ambience. In the latter case, closing the port may afford less exact control over the last of the extracted fluid from the column 15, although this may be less critical, for example, if the desired components of the sample to be extracted are suitably separated by density media, that are innocuous inclusions in the extracted fluid. Similarly, if port 12' leads to an off-chip vial, either a tubing connecting the chip to the vial, or the vial itself, is axis proximal of the fill level of the column 15, or the vial is a controlled, pressurized vessel, that is depressurized to admit the extracted fluid. Through the channel and hole 12', the extracted component of the fluid may be delivered to an off-chip vial or container that is carried by the centrifuge, mounted to the chip or carried by a pneumatic chip controller, or may be delivered to another layer of the chip, for example, for further analysis.

[0058] The outlet 16 has an axis-relative position that lies strictly between axis- proximal and axis distal ends of the column 15, and preferably lies within the region 14 which is used for determining alignment of components with the outlet 16. Preferably the outlet 16 lies between 1.1 times the axis-proximal radius (r _m in) of the chip, and 0.9 times the axis-distal radius (r _m ax) of the chip. The axis-proximal and -distal radii of the column spans at least 30% of an extent of the chip, typically 50-80%, and r _m in < ½ r _ma x. Accordingly there is a wide range of positions for the outlet that principally depend on the isopycnic surface relative to expected density gradients of the sample.

[0059] FIGs. 1A-E schematically illustrate principal steps in fraction extraction. FIG. 1A shows the sample 25, which for the purposes of exemplification may be whole blood, in the column 15 and extending into the J-shaped channel 19, as well as a high density medium 26 in chamber 17. As can be surmised by the curves defined by the liquids under centrifugation, the axis of the centrifuge is relatively close to a centre of top (axis proximal chip) edge along which the ports of the chip are arrayed, this being one common positioning of chips relative to the centrifugal axis when mounted via a chip controller, as it entails strong centrifugal gradient fields for a given chip length and centrifugation rate.

[0060] Note that outlet 16 is blocked throughout centrifugation, to prevent whole blood from entering the opening 12'. This may be accomplished with a suitable flow resistance, for example with a hydrodynamic resistance of the channel connecting the outlet 16 with the opening 12', with a valve, such as a syphon, or with a controlled positive pressure on opening 12' relative to column 15 from a corresponding port in an adjacent layer of the chip (not in view). At this juncture, each of ports 12a,b,c may be open to ambience. Indeed, as the fill lines of the sample 25 and medium 26 suggest, at least ports 12a, b are open to ambience, or were at some point since centrifugation began.

[0061] Blocking port 12b during loading of the sample, fractionation, and dispensing is one way to preclude sample from entering the J channel 19, which may be preferable for visualizing droplet injection, or for avoiding any loss of the extracted component. It is expected to be most efficient to maintain the port 12b open to ambience throughout the process.

[0062] FIG. 1 B shows a fractionated sample 25, which results from a high centrifugation rate for a sustained period of time. The fractionated sample 25 is shown to have 3 distinct discernible phases 25a, b,c, which will be named here plasma 25a, buffy coat 25b, and red blood cell (RBC) 25c although the sample could have substantially any number of phases (discernible or not), and could be of any expected composition. Note the buffy coat 25b is illustrated to be a disproportionately large volume fraction for human blood, to facilitate viewing. It will be assumed for this example, that the buffy coat 25b is the desired component (though this is not always the case for blood sampling), and the primary goal is purity of the extraction, as opposed to completeness of extraction, in that it is preferred to exclusively extract buffy coat 25b, and leave some buffy coat 25b in the column 15, rather than to extract all the buffy coat 25b with the minimum of plasma 25a and RBC 25c (or other density medium if used). The present invention is particularly useful when a desired part of the fractionation (which may or may not correspond with an entire discernible phase) has an a priori unknown, but bounded, fractional volume of the sample, or the desired component is precious, and losses due to overflow metering are undesirable, as both of these result in an unknown axis-relative position of a desired part.

[0063] The volume of column 15, and medium 26 and relative position of outlet 16 in comparison with the observed isopycnic surface is shown with a wide margin in that substantially less volumetric composition of plasma 25a could have been in the sample, and the chip would work. As will be appreciated, a simple trade off lies between greater accuracy of metering, and reduced inter-sample volumetric composition variations accounted for in the design, and a volume of the media required for the extraction process (both in the chamber 17 and in the column 15). The chip can be designed for any range of sample and medium volumes.

[0064] As was explained by WO 2014/111721 to Banks, cited above, remixing of fractionated sample is an on-going concern. Once the sample is fractionated to a desirable degree, the chip may be spun at lower rate to facilitate fluid dynamics with suitable efforts to avoid remixing. Such efforts include slow addition of the medium 26 into the column 16, and limiting mixing action within the column 16.

[0065] FIG. 1C shows the chip in a state as shown in FIG. 1 B after positive pressure is supplied to port 12c. Naturally it can equivalently be supplied by a negative pressure on port 12b (if the gas pressure at hole 12' is not blocked, but modulated, this may equally need to be depressurized by cooperative pressure supplies, and it may further be required or desired to limit draw through the J channel if the additional weight of the medium in supply chamber 17 doesn’t ensure dispensing before the sample rises to the opening 20, thus control may be appreciably simpler if positive pressure is applied at port 12c). First the medium 26 primes the serpentine channel 21 and then the liquid is dispensed drip-by-drip into the J-shaped channel 19. In other embodiments a colour, opacity, scattering, or index of refraction of the medium 26 is chosen for visual detection of an isopycnic surface, especially if the medium’s axis-relative position is useful for alignment, however in this case, the medium 26 has a density greater than that of the sample’s heaviest liquid component, and accordingly the only occasion to observe the medium would be within the J channel 19 which is not within the region 14 as shown. However, it will be appreciated that any observable phase boundary can be suitably used for determining axis-relative rise or fall of isopycnic surfaces within the column 25.

[0066] The droplets have sufficiently low affinity for walls of the J channel 19, and are small enough relative to the J channel, that they do not occlude the J channel, as this would trap air bubbles that would impair flow. To ensure this, a hydrodynamic diameter of the nozzle 22 may be, for example, smaller than 3/4 to 1/3 a cross-sectional area of the J channel. The droplets, under the centrifugal field, pass through the lighter, separated phases in the J channel 19. As the droplets fall through the occupying fractions, they tend to disturb the density separation, leading to a mixture that may more closely resemble whole blood than the fractions with phase boundaries prior to dispensing.

[0067] The medium 26 as illustrated has a density greater than any substantial phase of the fractionated sample, and thus the droplets first accumulate at a bottom of the J channel 19, until that channel is occupied by the medium 26. Once the bottom is filled, the droplets fall into the column 15 where they accumulate at the axis-distal end. Obviously the accumulation in the J channel 19 can be avoided by providing no local bottom of the J channel before meeting the column 15. FIG. 1 D shows a substantial volume of the medium 26 at the bottom of the column 15, which raises the buffy coat 25b towards the outlet 16. As the buffy coat 25b enters the region 14, an imaging system can monitor the rate of rise and modulate the pressure and centrifugation rates to automate the alignment of the desired isopycnic surface with the outlet 16.

[0068] FIG. 1 E shows the isopycnic surface separating the buffy coat 25b from the RBCs 25c just below the outlet 16. At this juncture, releasing or overcoming a hydrodynamic resistance on the extraction channel, for example by drawing negative pressure through opening 12', will draw the buffy coat 25b into an (on- or off-chip) extraction chamber (not in view). It may be equally critical to ensure that no plasma 25a is entrained into the extraction chamber. To this end, control can be applied by maintaining a hydrodynamic resistance on the extraction channel to limit flow rates and control the stoppage of flow when called for, as determined by the imaging system. Alternatively, the whole plasma 25a can be removed (with some of the buffy coat 25b) prior to the buffy coat 25b extraction, for example with a chip with bifurcated outlet 16 as shown in FIGs. 1 C',D',E'.

[0069] Herein variants of the embodiment of FIG. 1 bear the same reference numerals to denote functionally same or similar features, the descriptions of which are not repeated except to explain relevant differences. Herein each variation is presumed independent and combinable with each other variation to form further alternative variants and embodiments of the present invention, unless otherwise indicated.

[0070] FIGs. 1 C',D',E' show the same substrate 10 as shown in FIG. 1 , with one variation, the outlet 16 is bifurcated to extract multiple distinct components of the sample, one after the other. As such there are three openings 12'. The method could begin with extracting the plasma 25a, but this is not illustrated. What is shown is the extraction of the buffy coat 25b with an assumed density of the medium 26 being less than a densest of the phases of the fractionated sample 25, but denser than the extracted component. If the extracted component were denser than the medium 26, with a chip of this design, once the J channel 19 is filled with the medium 26, and lightest components of the sample, the medium 26 would rise within the column 15, stirring the contents thereof.

[0071] As shown in FIG. 1C', instead of the medium 26 accumulating at the bottom of the J channel 19, it accumulates at the phase boundary between the RBC 25c, and buffy coat 25b within the J channel. If it is necessary for a process to inject a lighter medium 26 into a fractionated column below a desired isopycnic surface, or to inject a denser medium 26 in to the column above the isopycnic surface, the chip can be again subjected to centrifugation to refractionate the mixture with the added medium 26. [0072] By FIG. 1 D', the medium 26 occupies a plug of the J channel 19, and displaces upwards the lighter phases in both the column 15 and J channel, however, none of the medium 26 has entered the column, and none will until a balance the weight with the content of column 15 and J channel displace the plug axis-distally to the (axis- distal) bottom of the J channel. The axis-relative position in the J-shaped channel 19 may be higher or lower than that of the column 15, depending on whether the medium 26 has higher or lower density than the average of the content of the column 15 (which varies moment by moment as it gains RBC 25c content). This variation is generally subtle, and is exaggerated in the illustration.

[0073] By FIG. 1 E', the volume of the medium 26 exceeds what can be held in balance in the J-shaped channel 19, and it flows into the column 15. While the medium 26 should stir the RBC 25c with the medium 26, leading to a cloudy mix (not amenable to illustration), the clear boundary between the fractionated buffy coat 25b should be maintained, and allow for monitoring by the imaging system. If a phase separates the mixed phases and the extracted component, or if the mixed phase is visibly discernible from the extracted component, the mixing of the phases may not impair alignment or extraction purity.

[0074] FIG. 2 is a schematic top plan view of a second variant of the substrate 10 of FIG. 1 , bearing several variations. The fractionation column 15 is a U shaped vessel, with a main column 15a, and a secondary column 15b that preferably has a smaller volume (for example as a result of a lower etch depth). The secondary column 15b is tapered, and has a cross-sectional area that grows axis-proximally: this allows for less sample to enter the secondary column initially, and allows for more volume to enter during medium addition. The main column has a pinched or waisted section in the vicinity of the region 14, and the outlet 16 goes to a droplet-fed mixing chamber 30 that is axis- proximal the column 15. Furthermore, the controlled delivery channel 18, though having the same general composition, feeds the medium near the axis-proximal end of the column 15. As such, the medium must have a density lower than that of the extracted channel, or if the medium of higher density is used, the sample will have to be refractionated after the medium is dispensed, as the sample will be remixed during transit therethrough. Refractionation is not preferred as it slows production, and requires higher accuracy dispensing with less observable information. The illustrated variant is designed for lower density medium, as the region 14 extends mostly above the outlet 16, and thus the extracted components start axis-proximal of the outlet 16, and added medium tends to shift the isopycnic surfaces distally, as is further described hereinbelow. [0075] The U shaped fractionation column 15, with the main column 15a, and a secondary column 15b has two free surfaces for the sample, and two vents, and therefore adding lighter medium via (nolonger quite so J-shaped) channel 19 results in (a smaller volume of) denser components of the main column 15a shifting to secondary column 15b, and thus the level in 15b rises, though not as much as 15a. Without this secondary column 15b, adding lighter medium won’t produce a change in elevation of extracted component 25b.

[0076] The main column 15a is waisted at and above the outlet 16. As illustrated in the prior art, waisted columns 15 substantially improve extraction control. One reason for this is the area of the isopycnic surface separating a desired from an undesired phase. Consider the two candidate desiderata: if the objective is to avoid entrainment of a denser phase while maximizing collection of the component, presentation of the component in a narrower channel reduces the volume of the component that is not extracted because it is too close to the heavier phase. If the objective is to collect all of the component with the least amount of the denser component, the narrower channel is better, as the minimum axis-relative position below the isopycnic surface required to ensure all of the desired component is collected entrains a smaller volume of the denser component. Decreasing the isopycnic surface area also decreases a length of the phase boundary, which ceteris paribus, makes imaging more difficult. As it is always possible to enlarge the image to a smaller field of view with suitable optics, this need not impair imaging.

[0077] A final notable difference between the chip’s second variant is how the extracted component is treated on-chip. The outlet 16 is coupled to droplet-fed mixing chamber 30, specifically at a substantially axis-equal position as a final level (l) of the content of main column 15a. As such, the fluid is drawn against centrifugal force by an over-bearing pneumatic force, such as provided by negative pressure at port 12f. Negative pressure at port 12f (relative to that of main column 15a) will draw the extracted fraction, after a priming delay, to a nozzle 32, which will dispense the extracted fraction as a sequence of droplets, into the mixing chamber 30. At the same time, and with a shorter priming delay, reagent from a vented reagent supply 33 chamber is drawn into the mixing chamber 30 via a nozzle 34, to inject a second droplet stream into the mixing chamber 30. This second droplet stream, under the centrifugal field, is drawn over top the nozzle 32 and cascades down over the sequence of droplets, that mix relatively quickly given their high surface area and proximities, according to the teachings of Applicant’s US 9,555,382. The converging streams of droplets do not dwell in the mixing chamber 30, but are directly dispensed into a readout chamber 35, such as a colorimetric, interface pinning chamber taught by Applicant’s co-pending WO/2021 /156844.

[0078] FIG. 2 schematically draws a few fill lines showing free surfaces of liquid contents of respective chambers and columns at respective time points. These fill lines show the axis position relative to the chip as mounting in a centrifuge (for which the chip was designed). For example, main and secondary columns 15a,b and medium chamber 17 have initial (I) and final (l) fill levels schematically shown. Furthermore, a band corresponding to an extracted component or buffy coat 25b is shown in three states (25b,b',b"), to show displacement of the desired fraction with increasing additions of the low density medium 26. Specifically buffy coat 25b is illustrated relative to initial fill level I, as whole blood might be fractionated (except that the buffy coat volume is exaggerated for visibility). Buffy coat 25' shows the location after some low density medium is added to the chamber. Note that this sent the buffy coat axis-distally in main column 15a, and axis-proximally in secondary column 15b as more RBC shifts to 15b. Within the waisted region of main column 15a, the buffy coat, can be seen to increase in axis-relative depth, as its surface area decreases. It will also be noted that, the region 14 shows a preferable feature: the whole buffy coat volume, including both axis-distal and axis-proximal interfaces, are within the region 14 throughout the whole process. This provides the most information for controlling droplet-based dispensing of the medium. In the secondary column 15b, the buffy coat’s axis-relative depth decreases as it move axis-proximally.

[0079] FIG. 2' schematically illustrates a variant of the process on the substrate 10 of FIG. 2 that is preferable if the buffy coat 25b is particularly precious. If it is unacceptable to lose the buffy coat that is trapped in secondary column 15b, prior to loading of the sample, a high density medium is injected to fill level l ₀ . As a dense medium, it will occupy the U bottom of the column 15. Thereafterthe sample is loaded, bringing the high density medium to level IIA, which is not quite enough to permit anything but the high density medium to occupy the secondary column 15b. Hence level h is the level of blood and high density medium in the column 15 before and after fractionation. Thereafter, introduction of low density medium via opening 20 leads to entrainment of RBC 25c in secondary column 15b, which may readily mix, but has no effect on the imaging of the phase boundaries within the fractionated sample. Increased added low density medium will continue this effect until the desired isopycnic surface is aligned with outlet 16.

[0080] While this is not illustrated in this figure as it is unnecessary, the substrate 10 can alternatively include a high density medium supply chamber for dispensing the high density medium into the column 15 prior to sample loading, e.g. via secondary column 15b. A high density medium supply chamber is generally uncalled for if the sample is loaded prior to centrifugation, as the benefit of preventing any whole blood from entering the secondary column 15b is only provided if the high density medium is supplied first, it may be preferable to provide a metered volume of the high density medium by the same loading system. That said, there are off-chip loading systems, such as Applicant’s WO 2020/100039 that facilitate automated off-chip loading during or in between centrifugation steps.

[0081] It will be noted that all of the features of the claimed chip need not be supplied on a single layer thereof. FIG. 3 is a schematic illustration of a third chip variant that is inherently incomplete in that several features, like a hydrodynamic resistance element coupling the medium chamber 17 and opening 20, and the coupling of the outlets 16 (which also has a different form) to readout chamber 35, are not provided on the layer shown. A collection of features can be supplied by connected layers of the chip.

[0082] Like the second variant, the third chip variant uses a lower density medium fed to the main column 15a axis-proximally. The shape of the column 15 is different, in that the secondary column 15b is coupled to an overflow chamber 15c that removes lower- density or mixed components when the secondary column 15b exceeds a volumetric threshold. This permits larger volumes of lighter density medium to be dispensed, and a greater variation in axis-relative position of the isopycnic surface relative to a volume of added lighter density medium.

[0083] The controlled delivery channel 18 in accordance with the third variant is substantially revised, and none of the ports 12a-d need be pressure controlled, and may serve only as loading ports and air vents. Instead of pressure controlled port 12c directing flow of the medium into the opening 20, centrifugal field persistently urges this flow, but this is blocked by a hydrodynamic flow resistance coupled between holes 36 located on an adjacent layer (not in view). This flow resistance may be a serpentine path that extends axis proximally of the fill level of medium chamber 17, or may be a pressure controlled channel. Active pressure is therefore required on ports of the adjacent layer to ensure droplet dispensing of the medium, and to selectively extract on outlet 16, and delivery of the extracted component (treated or not by operations on the adjacent layer) via hole 36 into the readout chamber 35.

[0084] The chip of the third variant has a multi-layer structure. For example, the substrate 10 shown may have relief patterning on one side, but the 4-hole outlet 16 and a few other holes 36 constitute open through holes, which operably meet channels and vented chambers, for example, of the adjacent layers. It may be preferable to supply all the ports of the chip in a single line on one edge of the chip from one side, to facilitate a single-clamp port coupling to all channels or valves of the chip controller.

[0085] The readout chamber 35 has 12 micro- or nano-structured functionalized wells, each of which having a respective target moiety for assaying the (treated or not) extracted component receive via hole 36. The readout chamber 35 is coupled to two additional chambers 33a, b, which are respectively coupled to ports 12e,f, for example, to supply a developer solution, and a cleaning solution to the readout chamber 35 (for example, as explained in Applicant’s copending WO/2021 /156844). To effect this, ports 12e,f may simply be closed vents that are opened to ambience to release the fluid in the chamber in one shot, for example as explained in FIG. 11 of Applicant’s US 10,702,868. Alternatively, further flow control devices could be provided from the same or other pressurized fluids supplies of the chip controller to control this fluid displacement. The readout chamber 35 has a syphon valve outlet for extracting all the fluid (except what is adhered to the wells) only once the readout chamber 35 is filled and the syphon valve is primed. The outlet leads to an edge port 39 for fluid removal without an on-chip waste reservoir, as taught in Applicant’s WO 2020/100039.

[0086] FIG. 3 shows fill levels in main column 15a, and medium chamber 17 offering a sense of where the axis is located. Unlike the previous embodiments, the axis is located right of center of the substrate. Column 15a shows that fill level 1 ( ), which is the level to which the sample fills the column 15 initially, is defined by the overflow level in secondary column 15b. The corresponding extracted component is shown as buffy coat 25b in both main and secondary columns. The overflow level in the secondary column 15b is the level at every stage until the fluid component is extracted from the column 15 through outlets 16. At level 2 (b) the main column 15a has risen incrementally (somewhat exaggerated), and buffy coat 25b' (separately) in main and secondary columns has shifted. Again, the buffy coat 25b' moves axis-distally in main column 15a, and axis-proximally in secondary column 15b. In some embodiments, the buffy coat 25b in secondary column 15b flows over before alignment with the outlets 16 is achieved.

[0087] FIGs. 4A-C are schematic illustrations of a fourth variant of the chip, in respective states of fraction extraction. The fourth variant has a waisted column 15, and a hydrodynamic resistance 19' coupling the column 15, axis-distally, to a vented metering chamber 44. The fill lines suggest that the chip is designed for mounting to a centrifuge with an axis-proximal edge of the chip aligned with the axis of the centrifuge. It will be appreciated by those skilled in the art that a relatively large number of chips can be mounted to a same centrifuge if the chips are mounted with this orientation, as opposed to orthogonal thereto as is more conventional. It will be appreciated that imaging thusly oriented chips requires a differently oriented camera, if all chips can be imaged, and typically only 6 or fewer can be mounted. Furthermore, cameras for thus mounted chips may require compensation for viewing an angle of incidence.

[0088] The controlled delivery channel 18 illustrated comprises the medium chamber 17 axis-distally coupled to a channel communicating with opening 20, but not with a hydrodynamic resistance (serpentine) channel. Rather a simple syphon channel 21" is used. With the opening 20 located axis-proximally of the medium chamber 17, centrifugal force will only raise the medium 26 in the channel to a level equal to the fill level of the chamber 17, and thus no medium is dispensed until a pressure in port 12c overbears the centrifugal force. Alternatively, the chip may be tilted, for example according to the teachings of Applicant’s WO 2015/181725 to dispense. No nozzle 22 is called for, as the medium 26 directly fills metering chamber 44 which may be to just beyond capacity. No precise control over the volume of medium dispensed through the opening 20 is needed because a level of the isopycnic surface is controlled after the medium 26 is supplied, in this embodiment.

[0089] FIG. 4A specifically shows the chip in a state after centrifugal fractionation. Some sample (though predominantly RBC 25c as shown) is drawn into hydrodynamic resistance 19' by centrifugal pressure alone during fractionating centrifugation. If the sample was loaded prior to centrifugation, and then set for fractionation speed, a race between fractionation and flow into the resistance 19' will ensue (assuming the ports 12c,b are open to ambience), and will ensure that primarily RBCs occupy the axis- distal bottom of metering chamber 44.

[0090] FIG. 4B shows the chip with the column 15 and metering chamber 44 filled. With positive pressure maintained at port 12c, medium 26 is conveyed through syphon 21", and flows into opening 20, where it falls under centrifugal force into metering chamber 44. At the instant imaged, the metering chamber has recently been filled, and is overflowing into overflow chamber 44a. This would be a moment when the pressure (or its equivalent) is stopped, as the medium 26 is being wasted.

[0091] A mass density of the medium is greater than RBCs 25c, as shown, though not necessary. The drawings may be somewhat schematic in this regard as the competing rates of sedimentation of the medium, mixing with the RBCs, and displacement through the resistance 19' may lead to a cloudy mixing chamber 44, a substantially pure medium in mixing chamber 44, or a full layer of the RBCs 25c floating on the top of mixing chamber 44. Any density greater than the isopycnic surface could be used, to the advantage of obviating refractionation after addition. The isopycnic surface in the column 15 rises proportionately with the addition of the medium 26, although a greater density within the metering chamber 44 results in the free surface being axis distal to that of the free surface in column 15. The hydrodynamic resistance 19' is principally used to ensure that flow into the fractionated column is gradual, to avoid remixing, and not principally to ensure metering. The control over the supplied volume of medium 26 by this manner of dispensing lacks accuracy, and the volume required is not known before fractionation, therefore the strategy invoked in this embodiment is to over fill the column 15, and then control a withdrawal of excess from column 15 with a better controlled process. Thus FIG. 4B shows overflowing of chamber 44. The delivery rate of the medium 26 is slower than the throughput of resistance 19' to ensure that there is no backup within the metering chamber 44 during the filling.

[0092] While the process illustrated shows RBCs 25 in the metering chamber 44, it will be appreciated that this can be avoided or reduced by loading the sample and dispensing the medium in tandem, or by supplying the medium first. Supplying the medium and loading the sample in tandem is particularly advantageous for minimizing how much of the column’s volume is occupied by the medium, while ensuring that no air pockets remain in the resistance 19' after fractionation. If it is challenging to ensure no air remains in the resistance 19' after fractionation, it is trivial to provide a bubble trap structure near the junction of column 15 for holding a volume of the hydrodynamic resistance 19' within the column 15 to preclude any bubbles from rising above the isopycnic surface (and thereby mixing) the fractionated sample. Such a structure may include a lip, and a pocket, and is conveniently positioned on a side-wall of the column 15 above the junction with hydrodynamic resistance 19'.

[0093] FIG. 4C shows the chip in a state nearly ready for extraction. Negative pressure has been applied at ports 12b,c to draw fluid from the column 15 since the state of FIG. 4B (although equivalently positive pressure or positive pressure and blocking at ports 12a, 12' have obviously the same effect). Once the metering chamber 44 is filled, it spills over into overflow chamber 44a, where this liquid, which can only contain the medium and possibly initial contents of chamber 44 in FIG. 4A, accumulates. The isopycnic surface in the column 15 can be monitored as the liquid is being extracted to slow a rate of extraction. While the image taken in FIG. 4C may look like it is in its final pose, as the isopycnic surface of buffy coat 25b is aligned with the outlet 16, however effecting control over hole 12' while maintaining the pressure at ports 12b, c may be technically challenging with most chip controllers, especially those with few independent pressurized fluid supplies. If the negative pressure was promptly stopped at the moment imaged, centrifugal force would rebalance the levels in the chamber 44 and column, leading to a rise of the isopycnic surface by a distance that varies with the mass distribution across the column and chamber. Thus the withdrawal continues until the alignment is, viewed dynamically, exceeded. If too much of the fluid is withdrawn, the withdrawal process stops, continued centrifugation will rebalance the free surfaces within the column and chamber 44, and the smaller volume within chamber 44 can be filled again, either returning to the state of FIG. 4B, or by injection of a smaller volume into chamber 44. The correct volume of medium can therefore be dispensed to align the isopycnic surface with the outlet 16, without prior knowledge of the sample constituency.

[0094] FIG. 5 is a schematic illustration of a fifth variant of the chip for centrifugal microfluidic fractionation and extraction. The fifth variant has no new features, but merely illustrates one collection of the differences in an embodiment of the invention. The J channel 19 meets the waisted chamber 15 at an axis-distal end thereof. The fifth variant has a substrate with at least one pressure controlled port 12c for dispensing high density medium 26 into the column 15 (positive pressure at port 12c, or negative pressure at port 12b), and an off-chip valve or other pressure control (not in view) for ensuring that extraction is limited to the desired component connected to opening 12'.

[0095] FIG. 5 illustrates how a pair of density media 46a, b can improve fraction extraction, essentially by substantially separating components of neighbouring densities. If a volume of homogeneous liquid, having a constant mass density intermediate the components of the sample, and being immiscible with the sample, is placed within column 15 prior to, during, or after, loading of the sample, during fractionation, each density medium 46 will essentially divide the fractionated samples in two. If the constant mass density of the medium 46 is chosen to align with a highest or lowest density of the component to be extracted, the medium 46 has particular advantages for 1- demarcating the isopycnic surface of interest, particularly if the medium 46 is chosen to have a colour, index of refraction, opacity, or other property to facilitate visual, optical, or electromagnetic inspection; and 2- ensuring that organic material that would otherwise be adjacent the isopycnic surface in the fractionated sample, does not enter the extracted component, as such organic material may confound or complicate analysis.

[0096] Applicant has found particularly that pairs of density media 46a, b that have close densities, efficiently permit extraction of components of narrow density variations, which can be highly desirable for simplifying analysis by providing an extracted component with reduced complexity and higher purity.

[0097] FIG. 5 schematically illustrates a fractionated chip with density media 46a, b flanking the buffy coat 25b. The density media 46a, b may be chosen to meet multiple constraints apart from its homogeneous mass density: low reactivity with all of the species within the sample; producing a visible marker at at least one of its specific isopycnic surfaces within the fractionated sample; readily separated from the extracted components if required for further analysis, by membrane, filter, pressure, thermal, evaporative, or chemical isolation; or having low optical background or interference with intended detection. Particularly recommended are mineral and hydrocarbon oils, cure- resistant polymer resins, and monomers, and mixtures, solutions and dilutions thereof.

[0098] The arrangement of the fractionated content of column 15 and J channel 19 is consistent with whole blood being loaded, followed by the media 46a, b, via port 12a, and then sustained high rate centrifugation. As is well known in the art, high rate centrifugation is necessary for fractionation of many samples. For example, blood, in a column with axis-proximal distance of 2 cm, and axis-distal distance of 7 cm, may have sufficient centrifugal force at centrifugation rates of about 5 Hz (typical low centrifugation rates) for most microfluidic processes such as valving and dispensing; 10-120 Hz may be required for some processes like bubble mixing, or to expedite resisted fluid movements; and centrifugation rates of 0.2-1 kHz are common for centrifugal fractionation. Centrifugation rates of 1 .2-2.5 kHz are typically associated with ultracentrifugation.

[0099] It should be noted that typical microfluidic centrifuges may not operate at rates above a few hundred Hz. In particular, centrifuges with rotary unions (i.e. revolute joints that permit fluid to be coupled between the stationary and rotary parts, also called “slip rings”), and particularly centrifuges with multi-channel fluid couplings, cannot cost- effectively be provided for operation at the rates desired for some fractionation processes. Contact-based electrical slip rings may decrease rate, and increase costs of high rate centrifuges, but wireless or contact-free slip rings, which operate by induction, are not limited by contact friction. Closed pressurized canisters, electromechanical valves, thermoelectric devices, electronic communications and control circuitry, and pneumatic supply conduits are lightweight additions to centrifuges that can allow any centrifuge with electrical supply to operate as a centrifugal chip controller. Furthermore, on-board pumps that may not operate during ultracentrifugation or higher rate operations, are generally not damaged by mounting to chip controllers and exposed to high accelerations. Accordingly on-board pumps can be provided that are activated only during lower rate processes. Finally, a chip can be centrifuged on a high rate centrifuge, or an ultracentrifuge, and then transferred to a centrifugal microfluidic chip controller on a lower rate centrifuge.

[0100] FIG. 6 is a schematic top plan view of a sixth variant of the chip for centrifugal microfluidic fractionation and extraction. The sixth variant is similarto the fifth variant, but for a plurality of pressure controlled ports 12d -5, and four chambers 48a-d, that are coupled with the chamber 17. Many of the ports may be coupled to electromechanical valves that simply block the port, or open them to ambient, and use an air plug to control dispensing, however port 12c5, which is used for bubble mixing, needs to have a positive pressure supply connectable thereto, which can also be used to pump the prepared medium into the J channel 19. This embodiment allows for the tailored formulation and testing of respective density media for isolating components.

[0101] As an example illustrating how the sixth variant may be used, Applicant hereby incorporates by reference, the all content of a paper entitled “On-the-fly Physical Property Changes of Aqueous Two-Phase Systems (ATPS) Using a Centrifugal Microfluidic Platform (CMP)”. This paper teaches how a centrifugal microfluidic chip controller with a chip having chambers 48 and 17 can be used for production of tiny sample volumes of tailored density medium. By bubble mixing (as taught in US 10,702,868) controlled volumes of water, a salt solution, polyethylene glycol (PEG) and dextran (DEX) or a polymer, an aqueous two-phase system (ATPS) can be formed. Above a critical concentration of the solution, a single-phase mixture is changed into a two-phase solution due to the thermodynamic incompatibility. Thus the ATPS may be a single phase until it mixes with the sample. After separation, distinct high and low density phases manifest.

[0102] The ability to produce and assay many different medium densities by iteration, to isolate desired, observable phase components of the sample, or with a labelled sample, is a cost-effective density determination procedure that can be performed with the sixth variant. As long as the component to be assayed can be viewed in the fractionated sample, and the medium itself can be viewed, one can readily compute or observe the medium density that produces an isopycnic interface. With an ATPS, one can seek to confine narrower and narrower density components within a sample, for example, to simplify chemical analysis.

[0103] Furthermore, with automation and machine-vision inspection of the region 14, and suitable controls over the ports 12, automated processes for density-based detection, medium generation, or extraction can be performed, on a chip, with tiny reagent volumes. [0104] FIG. 7 is a schematic illustration taken from US 10,702,868, showing a seventh variant centrifugal microfluidic fractionating chip mounted to a chip controller 50 integrated with a blade of a centrifuge. The seventh variant is essentially the embodiment of FIG. 1 but with an on-chip chamber 49 coupled to outlet 16 for holding the component extracted from the column 15. A respective port 12 is provided for controlled extraction of the component and a channel coupling the outlet 16 and chamber 49 may have a hydrodynamic resistance to improve control over extraction.

[0105] The four ports 12 of the chip are respectively coupled by tubing 51 (although integrated channels and a single clamp for coupling each port to a respective channel is preferred in some embodiments) to respective electro-mechanical valves 52 (only 3 identified for ease of illustration). Each valve 52 is operated by controller 55. The valves 52 are shown for coupling two ports to ambience, or a plenum 56, or closed, and two ports to the plenum 56, or closed. The plenum 56 is a shared source of pressurized, or depressurized gas selectively supplied to the ports 12 via the valves 52. The plenum 56 may be coupled to a pressurized fluid supply such as a pump 58 that is adapted to operate while the blade spins at low centrifugation rates, or a pressurized fluid supply, such as a canister. For larger volumes of gas, liquids or solids may be retained in the canister, and gas production may be controlled with temperature and/or mechanical pressure as required in the plenum 56.

[0106] In use, if the chip 10 is preloaded with sample and medium prior to mounting to the chip controller, centrifugation commences and the sample is fractionated, which may require all valves to be in a closed or vented state throughout, or simply not be actuated during fractionation. Once fractionation is completed, as evidenced by imaging of the column, for example, centrifugation rate may be slowed to microfluidic processing speeds. It will be noted that some samples may not require higher centrifugation rates for fractionation. The valves are operated to: block port 12 adjacent to the on-chip chamber 49, open to ambient the column and opening, and apply positive pressure to the medium chamber. The medium follows the serpentine path to the vented opening, and is discretized by the nozzle upon entry into the opening. The droplets sink under centrifugal force to the axis distal end of the column raising the isopycnic surface of the fractionated sample therein. The valve operating under positive pressure is then closed, or vented and alignment is assessed. If satisfactory, the valves are operated to: vent on-chip chamber 49; and pressurize all other ports, as this will draw fluid aligned with the outlet, into the on-chip chamber 49. Closing the ports ends extraction of the desired component. [0107] FIG. 7A schematically illustrates the chip and chip controller 50 mounted on axis 59 in an enclosed centrifuge chamber. A centrifuge 60 is shown with a motor casing driving chip controller 50. A lid of the enclosure 61 is provisioned with a camera 62 and strobing lights 63 (only two labelled) that permit imaging of the region 14 during low-rate centrifugation, and possibly even at high rate centrifugation (to assess completion of fractionation). The images captured by the camera, are fed, in real-time, to a processor, which may be, or is in communications with, controller 55. Preferably the processor analyzes the image data, and sends control signals to the controller 55, and the centrifuge to control the valves 52, and a rate of centrifugation. This allows a computer processor coupled to the imaging system and the controller 55 to predictively control an axis-relative position of an isopycnic surface relative to the outlet 16. With sufficiently accurate feedback, the control may be provided without calibration of the chip, its fluid contents, or the chip controller valves. Any architecture for the processing and analyzing imaging data can be used. The camera may further image a readout chamber 35 and thus the processor may ensure fractionation, alignment, extraction, and then read out.

Experiments

[0108] FIG. 8 is a schematic illustration of a chip used to demonstrate the present invention. The chip is a further variant of the present invention, distinguished in that the controlled delivery channel is similar to that of FIG. 4, with an overflow intended for like use, however the medium chamber (for high density medium) has a second hydrodynamic resistance for improved control over volume dispensation, and the hydrodynamic resistance is embedded in a J channel. The overflow 1 of the column permits the plasma to be removed and simplifies extraction by requiring a cut off only at the denser isopycnic surface of the extraction component.

[0109] FIG. 8A is a photograph of the chip as fabricated, with fractionated blood contained in the column, a plasma volume having spilled into the overflow 1 as a heavy oil was introduced below the RBCs. Some whole blood spilled into the overflow 1 during sample loading, and is now fractionated. At this juncture, the oil may be retracted into the metering chamber until a desired volume of the oil spills into overflow 2, to change axis- relative position of the isopycnic surface (of white blood cells) in the column.

[0110] This chip was fabricated by CNC machining the channels and reservoirs in a 100 mm x 50 mm ^c 6 mm thermoplastic part (Zeonor 1060R, Zeon Chemicals). After machining, bonding was achieved by bringing the machined part in contact with a bottom cover consisting of a 200 pm thick extruded thermoplastic elastomer layer (Mediprene OF 400M, Hexpol TPE) and a 125 pm thick polycarbonate film (McMaster-Carr). The assembled chip was then annealed at 50 ^° C for 12h in an oven to improve bonding strength. Two metal tubes were glued on the chip to provide connection to an external 2 mL vial using flexible tubing, as shown in FIG. 8A.

[0111] The chip was mounted to a chip controller that can supply controlled air pressure to the ports of the chip during centrifugation. The chip controller is provisioned with 8 independent ports that can be pneumatically connected to either the pressure provided by an air pump or to normal atmospheric pressure using a series of eight three-way valves. The top edge of the chip is placed at about 50 mm from the center of rotation. The controller is equipped with a flash and camera synchronized with the rotation using a trigger signal allowing imaging of the chip during centrifugation (one image per turn).

[0112] With this arrangement, the following process was performed. The chip was first filled with the following buffers: 550 pL of a density gradient medium (Ficoll paque plus, Sigma-Aldrich) in the Second medium reservoir and 500 pl_ of a heavy liquid (density of 2.85 g/cm ³ , LST heavy liquid, Central Chemical Consulting) in the High density medium reservoir. After filling, the loading port located on the Second medium chamber was blocked using adhesive tape. A volume of 1 mL of whole blood diluted 1 :1 with a phosphate-buffered saline solution was also added to the external 2 mL vial that was connected to the chip using tubing. The chip and the vial were secured on the rotating platform of the chip controller, appropriate counter weights were installed, and pneumatic manifold was connected to the system. Rotation speed of the rotating platform was then set to 800 rpm. The centrifugal force resulting from the rotation leads to the metering and transfer of about 480 pL of the density gradient medium to the Column (FIG. 9B). Excess volume flows to the Overflow 1 chamber. The pump of the chip controller is then started and set to provide an air pressure of 3 psi. Rotation speed is reduced to 600 rpm and port #1 is activated (i.e. connected to the pump) to transfer diluted blood sample to the Column and layer it on top of the density gradient medium (Fig. 9C). Note that port #1 is pneumatically connected to the external vial using tubing as shown in Fig. 8A. Excess blood sample transferred from the external vial flows to the waste Overflow 1 chamber. Pump is then stopped and port #1 is deactivated (i.e. connected to ambient atmospheric pressure). The blood sample was fractionated by applying 800 rpm centrifugation for 45 min (FIGs. 9D and 9E). Heavy liquid is transferred to the bottom of the Column by supplying pressure of 4 psi at port #3 (with a rotation speed of 800 rpm), resulting in a gradual increase in the level of the fractionated blood sample isopycnic lines (FIGs. 9F and 9G). As the transferred volume of heavy liquid increases, the top layer of the fractionated blood sample (i.e. plasma) is gradually transferred from the Colum chamber to the waste Overflow 1 chamber. Once the desired level is reached for the isopycnic surfaces, the pump of the chip controller is stopped and port #3 is deactivated (FIG. 9G). In the example provided in FIG. 9, the heavy liquid transfer was stopped when the buffy coat layer nearly reached at the top of the Column chamber. The liquid level in the Column was then lowered by supplying a pressure of 4 psi and activating simultaneously ports #4, 5 and 7. When the buffy coat layer nearly reached the level of the Column outlet channel (Fig. 9H), ports #4 and 5 were deactivated allowing transfer of the buffy coat to the WBC chamber (Fig. 9I).

[0113] FIGs. 9A-I are photographs of the column (region of interest) at different stages in the process defined hereinabove. Photograph A shows the empty column, as well as a candidate region of interest for pixel-based image analysis. Photograph B shows secondary medium present in the column. Photograph C shows whole blood added or “layered” over the secondary medium. Photographs D,E show the fractionation happening. Photographs F,G show the plasma being spilled into overflow 1 by addition of heavy oil medium axis-distally into the column. Photographs G,H show withdrawal of the heavy oil to shift the white blood cells/buffy coat axis distally, into alignment with the column outlet. Photograph I shows the column after the buffy coat is extracted.

[0114] While Applicant considers the photographs of FIG. 9 to clearly show how isopycnic surfaces can be followed during medium addition and removal according to the present invention, to illustrate the simplicity of machine vision based automation with low- cost machine vision processing, FIGs. 10A,B are provided. FIGs. 10A,B are two copies of photograph E of FIG. 9. FIG. 10A is the full-colour photograph, and FIG. 10B is the green layer only of the colour image, presented in grayscale. Stripping out red and blue colours emphasizes the isopycnic surface of the buffy coat (shown by arrow).

[0115] The ability to manipulate fluids with a pneumatic chip controller to effect fractionation of tiny samples has been demonstrated. Other advantages that are inherent to the structure are obvious to one skilled in the art. The embodiments are described herein illustratively and are not meant to limit the scope of the invention as claimed. Variations of the foregoing embodiments will be evident to a person of ordinary skill and are intended by the inventor to be encompassed by the following claims.